HAL Id: hal-02936881
https://hal.archives-ouvertes.fr/hal-02936881
Submitted on 11 Sep 2020
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Interest of hydrogeological modelling in the interpretation of pollen proxies for palaeo-climate
reconstitution: example from Senegalese niayes
Pascal Maugis, Lazar Aguiar, Christophe Grenier, Aurélien Claude, Raphaël Léger
To cite this version:
Pascal Maugis, Lazar Aguiar, Christophe Grenier, Aurélien Claude, Raphaël Léger. Interest of hydro-geological modelling in the interpretation of pollen proxies for palaeo-climate reconstitution: example from Senegalese niayes. Histoire climatique des déserts d’Afrique et d’Arabie, 2009, Paris, France. �hal-02936881�
Interest of hydrogeological modelling in the interpretation of pollen
proxies for palaeo-climate reconstitution: example from Senegalese niayes
Pascal Maugis
1
, Lazar Aguiar
2
, Christophe Grenier
1
, Aurélien Claude
3
, Raphaël Léger
4
➢
Complex trade-off exist between infiltration and sea level forcings. but
watertable level at Diogo is mainly controlled by recharge after 7.5 ky
BP. Depression depth, altitude and filling rate are local issues.
➢
The ~3.5 ky BP forest expansion expresses a climatic bonification, out
of phase compared to the rest of the sahelian Africa.
➢
The 8.2 ky BP dry event left no sign in Diogo niaye, but did in another
one, thus suggesting local specificities that need to be further
investigated.
1) Objectives and context
Niaye ecosystems shelter an azonal relic guinean vegetation under sahelian climate. Pollinic proxy records in cores (Fig. 1) show that a guinean closed
forest installed at 10500BP* during the African Humid Period (AHP), and thrived under a somewhat degraded state until 2500 BP. Sub-surface water flow
brings today the necessary water to these interdunal depressions, keeping some of them flooded, with a distribution linked to water stress (Fig. 5). But
the equilibrium was different in the past between underground water availability, surface run-off, varying regional hydrogeology (Fig. 2), sea level (Fig. 4)
and distance, and pluviometry (annual and seasonal). This study aims at:
explaining how gallery forest persisted through the mid-Holocene dry episode (7500-4000BP), though in a degraded state,
examining the dynamics of the end of the Tafolian humid pulsation (4000-3000BP), when vegetation turned to sahelian
weighting the relative strength of the different forcings, notably the transience of higher-scale hydrogeology (Fig. 2)
finally, examining the pertinence of a (sometimes assumed) distinctive relationship between observed pollens and regional climate.
* all datations are calendar
1 CEA Saclay DSM/LSCE
91191 Gif sur Yvette Cedex, France
2 UQAM/GEOTOP,
Montreal, Canada
3 LTHE, 4ONERA
2) Environmental context
3) Scenarios & Numerical features
contact : pascal.maugis@lsce.ipsl.fr
Table 1: Characteristics of the simulated periods. (1) sea level and distance to present-day
coast in front of Diogo. (2) “lacustrine” corresponds to bottom inundation, “palustrine” to
sporadic inundation or sub-surface groundwater, “channels” to active turbulent river,
“meanders” to soft flow rivers with sedimentation, “ponds” to stagnant water. (3) Altitude of
the niaye bottom relative to present at Diogo.(4) Min. and max. of inter-annual infiltration.
(5,6) Prescribed head, linearly interpolated between the nodes A-E (Fig. 2). (7) Calculated
head in the mesh containing Diogo. (8) Excess of groundwater level respective to
present-day H-H(0). (9) Difference in inundation depth [H-Z] - [H(0)-Z(0)].
Numerical modelling with Cast3M Finite Element Code.
Generalized darcy law, 3D EFMH, Picard algorithm.
BRGM and SRTM data.
Parameter fitting on stationary head data (1975).
F
ig
.
6:
N
um
er
ic
al
m
es
h a
nd
bo
un
d
ar
y co
nd
iti
on
s
4) Results
Watertable morphology
10000 BP (top): Heavy rains induce heavy underground flow diverging from the main
recharge area to the sea (west), to the Senegal Delta (north) and in-land (east). Seepage from the flanks of the Plateau de Thiès limits the increase of the watertable below, which stands deep below the ground along the western boundary, but close to the surface at the feet of the elevation.
7500 BP (center): Slightly less favourable sea-level and continental hydrogeology do not
contribute much to the degradation of groundwater availability. The abrupt occupation by ferns should therefore be interpreted as an effect of climate change. A climate as dry as today would lead approximately to present conditions (sahelian vegetation) and would thus be inconsistent with the observed vegetation. The climate shift must then have had a moderate amplitude, in terms of annual precipitation or of seasonal repartition.
2800 BP (not shown):The decline of Elais Guineensis and the increasing relative presence of
peri-aquatic species suggest aridification as soon as 3000BP, confirmed by Travi (1987). Hydrogeological conditions, very close to actual, indicate a climate change that spanned over several centuries.
Qualitative
forcing :
Dry periods
present P & ETR
Very dry
P/2, ETR/2
very humid
P*2, ETR*1.9
Platea u de T hiès recha rge ar eaFig. 1 : Synthesis of pollen analysis in
core Diogo II with known badly-dated
climatic events. Main features are (i)
forest expansion from 10500BP; (ii)
substantial degradation around 7500BP
with
fern
expansion
and
high
sedimentation rate; (iii) settlement of
humidity-loving Elais Guineensis
3000-4000BP; then (iv) shift to sahelian types
ca. 2500BP.
Fig. 4 : Relative sea level (without
surrection
correction)
during
quaternary. Ocean coastline was as
low as 120m below present during
Last Glacial Maximum. The measure
points correspond to different places
along west Africa coastline. Only
Faure worked on Senegal. Datations
are imprecise.
Fig. 5 : Repartition of vegetal species in the Niayes
(Raynal 1963, Michel et al. 1969). Depth to water
table is a primary forcing to its organisation.
Fig. 3 : Diogo niaye
system (blue), possibly
relic of a former stream
network. Diogo II core
location is plotted on the
local red zoom. Present
strong anthropisation is
clearly visible.
Afric an H umid Perio dHydrology in Diogo area
Calculated water table altitude along the Diogo transect (smoothed). Infiltration appears to control the overall level of groundwater. However, sea level and coast distance are the dominant factors locally. The influence of the eastern boundary condition remains beyond the central piezometric ridge but its effect is greater under low infiltration rates. Head gradient (and thus sub-surface fluxes) increase with infiltration, of course, but are limited by seapage when the watertable reaches the surface.
Late Hol ocen e Hum id p ulsa tion Peri-aquatic Peri-aquatic T ra n sv e rs al d une s T ra n sv e rs al d une s
?
C lim at C oa st al d une s C oa st al d une s Scenarios ResultsDepression Vegetation Climate
(m) indication (mm/y) (m) (m) (m)
11.5 -33 m -12.4 Dry [-3 ; 6 ] -20, -33 -40, -43, -36 -10.7 -18.8 -6.4
7 km Channels Very humid [-5 ; 14 ] -10, -28 +6.3 -1.8 +10.6
10 -23 m -10.0 Tropical Dry [-3 ; 6 ] -15, -23 -34, -34, -30 -6.6 -14.7 -4.7
5 km Channels humid forest Very humid [-5 ; 14 ] -7, -19 +8.9 +0.8 +10.8
7.5 -2 m -6.0 Idem Dry [-3 ; 6 ] -2 -21, -23, -10 +5.0 -3.1 +2.9
– ? more open Humid [-4 ; 10 ] -2 +10.7 +2.6 +8.6
Very humid [-5 ; 14 ] -2 +14.3 +6.2 +12.2
3.5 +1 m -4.0 Forest Very humid [-5 ; 14 ] +1 -4, -2, +4 +16.1 +8.0 +12.0
– Ponds expansion 2.5 +1 m -2.8 Forest Dry [-3 ; 6 ] +1 -3, +1, +5 +8.6 +0.5 +3.3 – Ponds decline 0 +0 m 0 Dry [-3 ; 6 ] +0 +1, +3, +10 +8.1 +0 – – Ponds Period (ky BP) Observation at Diogo Sea (1) Z (3) Recharge (4) H
coast A,B(5) Hland C, D, E (6) H (7) H' (8) ∆I (9)
type (2) Palustrine Periaquatic Palustrine Lacustrine Lacustrine Lacustrine Palustrine Periaquatic & sahelian
5) Conclusions :
Hydraulic Head (very humid) Hydraulic Head (humid) H(present) - H(7500 dry) 7500 B P (d eg rad ati o n ) Diogo ∘ 1m 2m 3m Hydraulic Head (m) Seepage location Watertable depth (m) 10 00 0 BP (A HP ) Diogo ∘Purely climatic interpretation of pollen data may lead to important errors in palaeo-climate reconstruction.
modelled zone
main recharge area
net evaporative balance
coa
sta
l du
nal
str
and
Diogo
city
Fig. 2 : hydrological features of the studied area.
(pink) bathymetry, (blue) regional piezometry
(dashes stand for altitudes lower than present sea
level).
Linguere depression, relic of last glacial maximum
low sea level, diverts much water from the main
recharge area situated on “Plateau de Thiès”
flanks. The regional hydrogeology is complicated
by drainage from/to meastrichtian confined aquifer
below, and by perched aquifers as well.
Hydrology at Diogo site
Calculated depth in the cell containing Diogo site. The model overlooks local fluctuations induced, for instance, by local topography, so differential piezometry H-H(0) and depression depth Z-Z(0) are used instead.
Impact of infiltration rate is clear. Under dry conditions, non-climatic forcings have and increased effect. And when combined , they have the potential to override the climatic forcing.
The delayed onset of vegetation may stem from ecosystem dynamics. The vegetation transition ~ 8.2-7.5 ky BP may be explained by minored infiltration. All humid scenarios yield around 10m inundation higher than at present, while all dry ones have it under 3m.These coherent results illustrates the important role of the depression depth, but are not always consistent with the observed sedimentology F os si l P ol le n D at a ba se (f ro m A .-M . L éz in e)